5700
J . Org. Chem. 1994,59,5700-5707
Is Through-Bond Dihydroaromaticity Attainable? Preparation of [4,5]Dihomotropone,Investigation of Its Ground-State Properties, and an Attempt To Generate the Dihomotropylium Cation Leo A. Paquette,* Timothy J. Watson, Dirk Friedrich, Roger Bishop,' and Eric Bacqub Evans Chemical Laboratories, The Ohio State University, Columbus, Ohio 43210 Received May 17, 1994@
Construction of the first dihomotropone (6) has been accomplished. The key elements of the synthesis were a regiocontrolled expansion of a l4.1.11 bicyclic a-bromo ketone and introduction of the dienone double bonds by means of the Garbisch method. The spectral properties of 6 under neutral and strongly acidic conditions are described and interpreted in terms of a lack of dihomoaromatic character. The parent bicyclol5.1. llnonadienyl cation is unstable and experiences ready Wagner-Meenvein rearrangement to the bicyclo[4.2.llnonadien-7-ylcation. These features appear well suited to the ultimate evaluation of through-bond interaction in tricyclic homologues of 6 that carry perpendicular n arrays in the third ring. Exploration of the myriad of options available for fostering interaction between nonconjugated n networks has preoccupied the attention of leading research groups for three decades. The search for homoconjugation, the result of through-space interaction across proximal, disjointed n systems was spearheaded by Winstein in the 1 9 6 0 ~Somewhat ~ later, pioneering work by several other investigators called attention to the remarkable fact that perpendicularly oriented n fragments linked together by a common tetrahedral carbon atom could also exhibit through-space interaction (spiroconjugation)under the proper circumstance^.^ In a seminal paper published in 1968, Hoffmann recognized that JC units also have the capacity for communicating electronically through u frameworks (throughbond interaction).* Gleiter in the late 1970s presented theoretical arguments showing that orthogonally oriented n networks should also find it possible to interact through the bonds of a cyclobutane ring, the consequences being opposite to those witnessed in spiroconjugated ens e m b l e ~ .The ~ existence of a relay orbital effect has since @Abstractpublished in Advance ACS Abstracts, August 15, 1994. (1)On sabbatical leave from the University of New South Wales, Kensington, New South Wales, Australia. (2)For reviews, see: (a) Winstein, S. Q.Rev. Chem. SOC.1969,23, 141.(b) Winstein, S. In Carbonium Ions; Olah, G. A., Schleyer, P. v. R., Eds.; Wiley: New York, 1972;Vol. 111. (c) Warner, P. M. Top. Nonbenzenoid Aromat. Chem. 1976,2. (d) Paquette, L. A. Angew. Chem., Int. Ed. Engl. 1978,17,106.(e) Childs, R. F.Acc. Chem. Res.
1984,17,347. (3)(a) Hoffmann, R.; Imamura, A.; Zeiss, G. D. J. Am. Chem. SOC. 1967,89,5215.(b) Simmons, H. E.; Fukunaga, T. J. Am. Chem. SOC. 1967,89,5208.(c) Gordon, M. D.; Fukunaga, T.; Simmons, H. E. J . Am. Chem. SOC.1976,98,8401. (d) Semmelhack, M. F.; Foos, J. S.; Katz, S. J. Am. Chem . Soc. 1972,94, 8637.(e) Semmelhack, M.F.; DeFranco, R. J. J . Am. Chem. SOC.1972,94,8838.(0Semmelhack, M. F.; DeFranco, R. J.; Margolin, Z.; Stock, J. J . Am. Chem. SOC.1973, 95,426.(g) Semmelhack, M. F.; Foos, J. S.; Katz, S. J. Am. Chem. SOC.1973,95,7325.(h)Boschi, R.; Dreiding, A. S.; Heilbronner, E. J . Am. Chem. SOC. 1970, 92, 123. (i) Batich, C.; Heilbronner, E.; Semmelhack, M. F.Helv. Chim. Acta 1973,56,2110.ti) Batich, C.; Heilbronner, E.; Rommel, E.; Semmelhack, M. F.;Foos, J. S. J . Am. Chem. SOC.1974,96,7662.(k)Tajiri, A.; Nakajima, T. Tetrahedron 1971,27,6089.(1) Dum, H.; Gleiter, R. Angew. Chem., Int. Ed. Engl. 1978,17,559.(m) Gleiter, R. Top. Curr. Chem. 1979,86,197. (4)(a) Hoffmann, R.; Imamura, A.; Hehre, W. J. J . Am. Chem. SOC. 1968,90,1499. Reviews: (b) Hoffmann, R. ACC.Chem. Res. 1971,4,1. See also: (c) Gleiter, R. Angew. Chem., Int. Ed. Engl. 1974,13,696. (d) Paddon-Row, M. N. ACC.Chem. Res. 1982,15,245. (e) Paddon-Row, M. N.; Jordan, K. D. In Modern Models of Bonding and Delocalization; Liebmann, J. F., Greenberg, A., Eds.; Verlag Chemie: Weinheim, 1988; p 115.(0 Martin, H. D.; Mayer, B. Angew. Chem., Jnt. Ed. Engl. 1983,
22,283.
been substantiated in these laboratories6 and those of Gleiter.' Thus, 2-fold annulation of a cyclobutane core across its 1,3-and 2,4-positions as in 1-3 serves to lock the individual n ribbons into a perpendicular relationship. Stabilization is intrinsically linked to the number of n electrons involved and materializes when the total count is 4n,but not (4n 2).
+
a*a 1
2
3
Since the method of choice for assessing the level of interaction present in neutral hydrocarbons such as 1-3 is photoelectron spectroscopy (PE),* one is necessarily dealing with the respective radical cations. Consequently, charge dispersal must be factored into the already complicated interplay of bonding and antibonding electronic levels. For this reason, we have been attracted to seeking out ground-state effects that could rival those a t play within the confines of a PE spectrometer. Much in the way that tropone is generally regarded to be a representative nonbenzenoid aromatic s y ~ t e m ketone ,~ 4 or 5 (but unlikely both) could exhibit enhanced polarization because the prevailing orbital exchange contributions are well suited to long-range interaction through the Walsh orbitals of the four-membered ring. ( 5 ) (a) Bischof, P.; Gleiter, R.; Haider, R. Angew. Chem., Int. Ed. (b) Bischof, P.; Gleiter, R.; Haider, R. J. Am. Chem. Engl. 1977,16,110. SOC.1978,100,1036.(c) Gleiter, R.; Schafer, W. ACC.Chem. Res. 1990,
23,369. (6)(a)Paquette, L.A.; Dressel, J.; Chasey, K. L. J.Am. Chem. SOC. 1986,108,512. (b) Dressel, J.; Paquette, L. A. J.Am. Chem. SOC.1987, 107,2857.(c) Paquette, L.A.; Dressel, J.; Pansegrau, P. D. Tetrahedron Lett. 1987,28,4965.(d) Dressel, J.; Chasey, K. L.; Paquette, L. A. J. Am. Chem. SOC.1988,110, 5479.(e) Gleiter, R.;Toyota, A.; Bischof, P.; Krennrich, G.; Dressel, J.;Pansegrau, P. D.; Paquette, L. A. J.Am. (0 Dressel, J.; Pansegrau, P. D.; Paquette, Chem. SOC.1988,110,5490. L. A. J . Org. Chem. 1988,53,3996.(g) Paquette, L.A.; Dressel, J. In Strain and Its Implications; de Meijere, A., Blechert, S., Eds.; Kluwer Academic Publishing: 1989;pp 77-107. (7)(a)Gleiter, R.; Sander, W.; Butler-Ransohoff, I. Helv. Chim. Acta 1986,69,1872.(b) Gleiter, R.; Muller, G.; Huber-Patz, U.; Rodewald, H.; Irngartinger, H. Tetrahedron Lett. 1987,28,1985.(c) Gleiter, R.; Steuerle, U. Tetrahedron Lett. 1987,28,6159.(d) Gleiter, R.; Muller, G. J . Org. Chem. 1988,53,3912.(e) Gleiter, R.;Steuerle, U. Chem. Ber. 1989,122,2193. (8)Turner, D. W.; Baker, A. D.; Baker, C.; Brundle, C. R. Molecular Photoelectron Spectroscopy; Wiley: New York, 1970.
0022-326319411959-5700$04.50/00 1994 American Chemical Society
J. Org. Chem., Vol. 59,No. 19, 1994 5701
Preparation of [4,5]Dihomotropone
Scheme 1
4
5
As a prelude to this effort, we describe herein the preparation of 6,the key structural component of 4 and 5.1° The spectral properties of this cross-conjugated ketone reveal that interruption of the tropone x network with a pair of methano bridges does not give rise to a dihomoaromatic entity.l1 Consequently, the absence of measurable ground-state perturbations in 6 qualify it as a n entirely suitable constituent for our eventual examination of 4 and 5. The availability of 6 has also provided us with an opportunity to examine whether the related cations 7a and 7b are entirely classical or gain stability on the basis of charge delocalization into the cyclobutane Walsh orbitals as implied by the dashed lines in 8.lZUnfortunately, the rapidity with which 7a and 7b experience structural isomerization has precluded experimental evaluation of their electronic character.
10
9
a/
/
0 HO H 0
3
"
o
-
78, A = OH
b,R=H
a
0 13
12
KOH, H20; Hfi04. SOCI2, C6H6 CH30H. Ba(OH)2,A.
'
CH2N2. dAgOAC, Et3N,
Scheme 2
15
14
6
11
16
8a,R=OH b,R=H
Results Synthetic Considerations. Widely different pathways are potentially available for conversion of the known cyclobutane diester 913into dienone 6, and several of these were investigated. One option considered to be quite direct involves homologation of both functionalized chains in 9 so as to produce symmetrical diester 10, followed by Dieckmann cyclization as a means of generating keto ester 11 (Scheme 1). While application of the h d t - E i s t e r t sequence to 9 proved entirely workable, various attempts to achieve ring closure even under high dilution conditions14were to no avail. Unreacted 10 was returned in every instance, presumably because the preferred diequatorial status of both propionate residues serves as a powerful deterrent to ring closure. A similar fate accompanied the heating of diacid 12 with barium hydroxide up to 260 "C at atmospheric pressure and (9) (a) Streitwieser, A., Jr. Molecular Orbital Theory for Organic Chemists; John Wiley & Sons, Inc.: New York, 1961; p 279. (b) Nozoe, T. In Non-Benzenoid Aromatic Compounds; Ginsburg, D., Ed.; Interscience Publishers: New York, 1959; Chapter VII. (c) Ferguson, L.N. The Modern Structural Theory of Organic Chemistry; Prentice-Hall, Inc.: Englewood Cliffs, NJ, 1963; p 383. (d) Lloyd, D. CarbocyclicNonBenzenoid Aromatic Compounds;Elsevier Publishing Co.: Amsterdam, The Netherlands, 1966; Chapter VI. (e) Badger, G. M. Aromatic Character and Aromaticity; Cambridge University Press: London, 1969; p 85. (0For a contrary opinion, see: Bertelli, D. J.; Andrews, T. G., Jr. J.Am. Chem. SOC.1969,91, 5280. Bertelli, D. J .;Andrews, T. G., Jr.; Crews, P. 0. J. Am. Chem. SOC.1%9,91, 5286. (10) Preliminary communication: Paquette, L. A.; Bacqu6, E.; Bishop, R.; Watson, T. J. Tetrahedron Lett. 1992, 33, 6559. (11)Haddon, R. C. Tetrahedron Lett. 1974, 2797. (12) Strictly speaking, the Haddon model of the dihomotropylium cation is based on a bicyclo[l.l.0lbutane ring as the mediator of electronic interaction. Species 7 and 8 are therefore dihydro analogues of Haddon's proposed network. Since "din is the "only suitable prefix in the lexicon," we adopt it in the present context, but prefix the word with the through-bond connotation. (13) Gleiter, R.; Bischof, P.; Volz, W. E.; Paquette, L. A. J. Am. Chem. SOC.1977, 99, 8. (14)Ewing, G. D.; Paquette, L. A. J. Org. Chem. 1975,40, 2965.
19
ia
17
*
aCH& (5 eq), (C2H5)2Zn, Et20,35 OC. FeC13, DMF, ti. KOAc, TsS(CH2)3STs, EtOH, reflux. dLIAIH4, THF.
under reduced pressure.15 These conditions led to tarry polymer formation. In light of these developments, the decision was made to capitalize on the efficiency with which 9 undergoes the acyloin reaction to give 14.13Simmons-Smith cyclopropanation16 of this intermediate and direct Saegusa oxidation17 of 15 successfully gave rise to 16 in a n optimized overall yield of 55% (Scheme 2). At this juncture, the &diketone underwent smooth condensation with trimethylene dithiotosylate'* as long as the usual strongly basic promoterslg were replaced with potassium acetate in refluxing ethanol.20 Since 17 proved unreactive to diisobutylaluminum hydride, reduction to diol 18 was effected with LiAlH4. In addition to 18, two other products were formed, neither of which was the trans diol. These byproducts were identified as 20 and 21. The overreduction likely arises via retroaldol cleavage of the monoreduced intermediate 22 and further hydride attack a t the aldehyde carbonyl unmasked in this manner. The a-keto dithiane part structure presumably exists as a n enolate while the reduction proceeds and is therefore (15)Thorpe, J. F.; Kon, G. A. R. Organic Syntheses; Wiley: New York, 1932; Collect. Vol. 1, p 192. (16) Simmons, H. E.; Cairns, T. L.; Vladuchick, S.A.; Hoiness, C. M. Org.React. 1973,20, 1. (17) (a) Ito, Y.; Sugaya, T.; Nakatsuka,M.; Saegusa, T. J.Am. Chem. SOC.1977, 99, 8366. (b) Ito, Y.;Saegusa, T. J. Org. Chem. 1977, 42, 2326. (c) Ito, Y.; Nakatsuka, M.; Kawamoto, F.; Saegusa, T. Org.Synth. 1980, 59, 113. (18) Woodward, R. B.; Pachter, I. J.; Scheinbaum, M. L. Org.Synth. 1974, 54, 33. (19) Bryant, R. J.; McDonald, E. Tetrahedron Lett. 1975,16, 3841. (20) Wakamatsu, T.; Akasaka, K.; Ban, Y. J. Org. Chem. 1979,44, 2008.
6702 J. Org. Chem., Vol. 59, No. 19, 1994
Paquette et al.
20
21
a H
(21)Bacque, E.Unpublished observations. (22)Dave, V.; Warnhoff, E. W. J. Org. Chem. 1983,48,2590. (23)Garbisch, E.J. Org. Chem. 1965,30,2109. (24)For comparison, see: Chasey, K. L.; Paquette, L. A.; Blount, J. F. J . Org. Chem. 1982,47,5262.
b
MsO 23
22
The route that ultimately proved workable began by hydrolysis of 14 to a-hydroxy ketone 23 as previously described.13 The conversion of this readily available intermediate into the bromide via mesylate 24 proved uneventful (Scheme 3). Although enone 26 could be produced from the mesylate, this attractive compound could not be enticed into further useful chemistry.21 Similarly, the direct ring expansion of the a-bromo ketone t o 26 with diazomethaneZ2was not adequately regioselective for our purposes. In contrast, recourse instead to the lesser reactive ethyl diazoacetate in the presence of boron trifluoride etherateZ2 induced smooth ring expansion to give only 27 (88% isolated yield). As anticipated, the latter was obtained as a keto-enol mixture of both possible diastereomers. Reductive debromination of 27 with zinc dust in acetic acid and subsequent heating of the resultant P-keto ester in 5 M HC1 proved highly suited to acquisition of the tetrahydro precursor 13. The CzUsymmetry of this ketone was evident from its simplified lH N M R and fiveline 13C NMR spectra. In order to apply the Garbisch protoc01,2~13 was converted to its ethylene ketal and reacted in turn with 2 equiv of bromine in ethylene glycol as solvent. The trans disposition of the two bromines in the exclusive product of this very clean halogenation was likewise immediately recognized on the basis of its 13C NMR spectrum. The observed six lines require that a CZaxis be present. The C,-symmetric cis isomer would be represented instead by eight carbon signals. The 2-fold dehydrobromination of 28 with potassium tert-butoxide in warm DMSO was also achieved without complication. Hydrolytic removal of the ethylenedioxy group had necessarily to be accomplished rapidly under conditions as mildly acidic as possible because of the acid sensitivity of 6 (see below). Brief shaking (5 min) with 3% aqueous sulfuric acid emerged as a reasonable compromise. Spectral Characteristics of 6. The infrared carbonyl stretching frequency of 6 in CCl4 solution is located at 1653 cm-', a somewhat higher wavenumber relative t o the values reported for 2,6-~ycloheptadienone(1647 cm-l) and 2,7-~yclooctadienone(1640 ~ m - ' ) . ~By~ comparison, the carbonyl absorption of tropone makes its appearance at 1597 cm-l; this lowering has been construed as indicative of contributions from the polarized form. The modest gravitation of 6 in the opposite direction quite probably arises from the greater conformational strain present in the bicyclo[5.1. llnonadienone framework.24 The same trend is evident in the stretching
:a
Scheme 3
insulated from more advanced reduction. For this reason, the conversion of 20 to 21 is considered to take place during the hydrolytic workup. Unexpectedly, all attempts to effect the dehydration of 18 did not lead to 19.21 Nor were experiments aimed at unmasking the latent carbonyl in 18 successful.
24
26
25
13
27
20 e CH~SOPCI, py, CH2Cl2, 0 'C. d
6
* LiBr, Li2CO3, DMSO, 70-100 'C.
'LiBr, acetone, reflux. CH2N2, CH30H. * N2CHCOOE1, BFyOEt2, CH2Cl2. 'Zn, HOAc, Et20. '5M HCI, acetone, reflux. HOCH~CH~OH,, (TSOH), CgHg, reflux. Br2 (2 equiv), HOCH2CHzOI-I. KOf-Bu, DMSO, 50 "C. k3% H2SO4, shaking 5 min.
*
'
'
frequencies of the double bonds in these conjugated dienones. This absorption, which appears a t 1623 cm-' in 6, is displaced by 10 wavenumbers relative to that at 1613 cm-l seen in both simpler structural analogues. The dihomotropone is characterized by a single ultraviolet absorption maximum in ethanol a t 245 nm (log E 3.71). This electronic behavior mirrors that exhibited by 29 and is bracketed by the two bands reported for 30.26 A comparison of these data with the electronic spectrum of 4,5-homotropone ( 3 V 6 is informative. We conclude that these experimental determinations provide no hint of charge separation in the ground state of 6.
29
XK,243.5 (loge 4.03)
31
30 266 (log E 3.40) 235 (log e 4.03)
32
kFgH290 sh (log e 3.66) 264 (log E 3.88)
The NMR spectral parameters for 29 and 30 should model well the olefinic proton chemical shiRs of 6 should it lack dipolar character. As seen in Table 1, the similarities between 6 and 30 in CDCls solution are striking. The overlapping of the a and ,6 protons observed for 29 can be attributed to those conformational readjustments required to bring this dienone to its global energy minimum geometry. Notwithstanding, none of the above (25)Krabbenhoft, H.0.J. Org. Chem. 1979,44,4285. (26)Chapman, 0.L.;Fugiel, R. A. J.Am. Chem. SOC.1969,91,215.
Preparation of [4,5]Dihomotropone
J . Org. Chem., Vol. 59, No. 19,1994 6703
Table 1. High Field 'H and
6 6.61(dd, 2 H) 6.00(d, 2 H) 3.01(m, 2 H) 2.88(m, 2 H)exo 1.82(m, 2 H)endo 194.98 146.78 130.70 31.68(d) 30.09(t)
lH NMR (300MHz)
13CNMR (75MHz)
a
N M R Data for Selected Unsaturated Medium-RingKetones (CDCk Solutions; Values in ppm)
1SC
29 6.10(m, 4 HP 2.25(m, 4 H) 1.70(m, 2 H) 193.23a 141.59 135.93 27.14 25.04
30 6.60 (m, 2 H)" 6.08(d, 2 H) 2.45 (m, 4 H)
32 6.82 (m, 2 H) 6.67 (m, 4 H)
192.50" 144.33 133.50 27.30
187.35 141.28 135.60 134.15
Reference 25.
Table 2. Compilation of Chemical Shift Data for 6,29,30,and 32 in Acidic Media (300and 75 MHz; Values in ppm)
8 8.39 (m, 2 HIa 6.75(m, 2 H) 3.65(m, 2 H) 3.38 (m, 2 H)exo 1.59 (m, 2 H)endo 7.13 (dd, 2 HF 6.28(d, 2 H) 3.26 (m, 2 H) 3.09 (m, 2 H)ex0 1.90(m, 2 H)endo
'H NMR FSO3H PPm
lH NMR CF3C02D PPm
13CNMR CF3COzD
e
158.23d 130.22 34.12 (d) 30.28 (t)
PPm
29 7.98 (dt, 2 H)b 6.95 (d, 2 H) 2.46 (m, 6 H)
8.64(m, 2 H)d 8.48 (dd, 2 H) 8.39(d, 2 H)
199.46d 150.41 135.75 30.30 27.15
184.7od 152.49 148.06 141.30
"C. Reference 27. Measurement made at 25 "C.e Signal too weak to be
dienones feature a n induced ring current suggested by the rather dramatic deshielding exhibited by tropone
(32). The protonation studies carried out on 6 were more limited than usual because of the inherent sensitivity of this dienone to acid. In FSOsH a t -78 "C, however, it was possible to observe the formation of 7a and to record the lH NMR (but not 13C NMR) spectrum of this cation prior to the onset of decomposition. As seen in Table 2, the protonation of 6 is accompanied by downfield shifting of the a-vinylic (Ad -1.78 ppm), p-vinylic (Ad -0.75 ppm), bridgehead (Ad -0.64 ppm), and exo-methylene protons (Ad -0.50 ppm) but not the endo-methylene protons which in contrast experience modest shielding (Ad +0.23 ppm). The chemical shift displacements observed for the lateral protons find close analogy in the effects encountered following comparable protonation of 29 and 30.27 As a consequence of the close structural similarities of these reference cations, e.g. 33, the comparison is considered to be direct and not beset with gross assumptions.
*
....'*
q
*. .
..a'
OH
OH
33
34
The usual diagnostic of homoaromatic character is the chemical shift difference observed for those protons positioned on the bridging carbon.2,28 Notwithstanding the lack of detailed knowledge of the origin of ring current effectsz9and possible contributions from local anisotropic influences, large Ad values for exdendo proton pairs have (27)Noyori, R.;Ohnishi,Y.; Kat8, M. J. Am. Chem. SOC.1972,94, 5105.
32
6.88 (dt, 2 HId 6.57(d, 2 H) 2.47(m, 4 H) 2.00 (m, 2 H)
a Measurement made a t -78 "C. Measurement made a t 0 observed because of ongoing decomposition.
Q
30 8.17 (br d, 2 HF 7.05 (d, 2 H) 3.17(br s, 4 H)
rather indiscriminately been interpreted in terms of induced diamagnetic ring current effects.30 X-ray crystallographic studies of monosubstituted homotropylium ions by Childs and co-workers31 have confirmed t h a t cyclic charge delocalizati~n~~ does indeed operate in these systems. Consequently, the very large Ad for the C-8 protons in the parent homotropylium cation (5.86)33can unquestionably be attributed in part to homoconjugation.
However, local anisotropic contributions are now recog(28)(a) Rosenberg, J. L., Jr.; Mahler, J. E.; Pettit, R. J. Am. Chem. SOC.1962,84,2842. (b) Winstein, S.;Kreiter, C. G.; Brauman, J. I. J. Am. Chem. Soc. 1966,88,2047. (c) Keller, C. E.; Pettit, R. J. Am. Chem. Soc. 1966, 88, 606.(d) Winstein, S.; Kaesz, H. D.; Kreiter, C. G.; Friedrich, E. C. J. Am. Chem. SOC.1965,87,3267.(e) Warner, P.; Harris, D. L.; Bradley, C. H.; Winstein, S. Tetrahedron Lett. 1970,4013. (0 Paquette, L.A.; Broadhurst, M. J.;Warner, P.; Olah, G. A.; Liang, G. J. Am. Chem. Soc. 1973,95,3386.(g) Olah, G. A.; Staral, J. S.; Liang, G. J.Am. Chem. SOC.1974,96,6233. (h) Olah, G. A.; Staral, J. S.;Spear, R. J.; Liang, G. J. Am. Chem. SOC.1975,97,5489.(i) Oth, J. F. M.; Smith, D. M.; Prange, U.; Schroder, G. Angew. Chem., Int. Ed. Engl. 1973,12,327. (29)(a) Johnson, C. E., Jr.; Bovey, F. A. J . Chem. Phys. 1958,29, 1012.(b) Haigh, C. W.; Mallion, R. B. Org. Mugn. Reson. 1972,4,203. (c) Agarawal, A.;Barnes, J. A.; Fletcher, J. L.; McGlinchey, M. J.; Sayer, B. G. Can. J. Chem. 1977,55,2575. (30)For other attempts to define homoaromaticity by experiment, see: (a) Dauben, H. J., Jr.; Wilson, J. D.; Laity, J. L. Nonbenzenoid Aromat. 1971,2, 167.(b) Feldman, M.; Flythe, W. C. J. Am. Chem. SOC.1971,93,1547.(c) Stam, C. H.; Counotte-Potman,A. D.; van der (d) Hoskin, D.H.; Wooden, Plas, H. C. J. Org. Chem. 1982,47,2856. G. P.; Olofson, R. J . Org. Chem. 1982,47, 2858. (e) Childs, R. F.; Mulholland, D. L.; Varadarajan, A.; Yeroushalmi, S. J. Org. Chem. 1983,48,1431.
(31)(a) Childs, R. F.; Varadarajan, A.; Lock, C. J . L.; Faggiani, R.; Fyfe, C. A.; Wasylishen, R. E. J . Am. Chem. SOC.1982,104,2452.(b) Childs, R.F.; Faggiani, R.; Lock,C. J. L.; Mahendran, M. J . Am. Chem.
1986,108,3613. (32)For recent theoretical treatments of this question, consult: (a) Haddon, R. C. J . Am. Chem. SOC.1988,110,1108.(b) Cremer, D.; Reichel, F.; Kraka, E. J.Am. Chem. Soc. 1991,113,9459and relevant
Soc.
references cited therein. (33)Warner, P.; Harris, D. L.; Bradley, C. H.; Winstein, S. Tetruhedron Lett. 1970,4013.
Paquette et al.
5704 J . Org. Chem., Vol. 59, No. 19, 1994 Table 3. '€ NMR I Analysis of the Cyclopentene and 1,3-CycloheptadieneEffects in Bisannulated Cyclobutanes
/H Y
/H
$2
X
x$
Scheme 4 6 -
a
/H
/H x*
H 35
x*
36a, R = Ac b.R=H
H'
-.I.,
c
3.53
1.69
1.27
3.76
"05
H
-I
3.45
.._.
2.40
37a, R = Ac b,R=H
A6 -1.31 1.27
A8 -1.80
_._.
+0.04
1.17
A8 -1.91
1.05
1.23
AS
bs -1.04
2.23 a NaBH4, CeCI3-7H20, CH3OH.
AczO, HOAc, (TsOH). NaOH, C H 3 0 H , H 2 0 . d M n 0 2 , CDC13.
A6 -0.96 2.48
2.11
2.01
as +0.10 .._.
1.24
1.44
A8 -0.20
H-l 2.42
3.54
A8 -1.12
38
2.55
2.48
A6 +0.07
nized to account for more than 40% of the total chemical shift d i f f e r e n ~ e . ~ ~ In the present context, the need exists to call attention to the striking interplay of local magnetic anisotropy effects and chemical shift that operates in 1,3:2,4bisannulated cyclobutanes (Table 3).@a5 Several interpretations have been accorded to the considerable deshielding that can accompany the placement of a methine proton above a region of unsaturation. Anisotropic field effects, differing degrees of steric compression, varying strain contributions, and orbital interactions through the filled cyclobutane orbitals have all been implicated. While these interpretations must currently be regarded as speculative, it is clear that wide variations in chemical shift often accompany structural modifications within molecules of this general type. The magnitude of the shiR differences between the ex0 and endo protons in 6 (Ad 1.06) and 7a in FS03H solution (Ad 1.79) are considered t o be a reflection of the greater delocalization that accompanies conversion to a pentadienyl cation subunit. The greater spread in Ad for 7a does not appear to be indicative of dihomoaromatic status for this species. This is tantamount to saying that charge density is not being propagated into the four-membered ring in the manner implicated in structure 8a. This conclusion is supported by the virtually identical chemical shift of the methylene carbons in 6 (30.09 ppm) and its protonated ion (30.28 ppm). The various pieces of evidence, when taken together, indicate that the cation in question is classical, although no good model system against which to compare 8a exists at present. In CFZCOZD solution a t 25 "C, it was possible to observe 7a for a more prolonged period of time. The lH (34) Childs, R. F.; McGlinchey, M. J.;Varadarajan, A. J . Am. Chem. SOC.1984,106,5974. (35)(a) Christl, M.; Herzog, C. Chem. Ber. 1986, 119, 3067. (b) Christl, M.; Herzog, C.; Kenner, P. Chem. Ber. 1986, 119, 3045. (c) Gleiter, R.; Zimmerman, H.; Sander, W.; Hauck, M. J. Org. Chem. 1987,52,2644. (d) Christl, M.; Herzog, C.;Nusser, R. Chem. Ber. 1986, 119, 3059. (e) Christl, M.; Herbert, R. Org. Mugn. Reson. 1979, 12, 150. (0 Christl, M; Bruntrup, G. Chem. Ber. 1974, 107, 3908. (g) Wilzbach, K. E.; Ritscher, J. S.; Kaplan, L. J . Am. Chem. SOC.1967, 89,1030. (h) Meinwald, J.; Kaplan, B. E. J . Am. Chem. SOC.1967,89, 2611. (i) Sander, W.; Gleiter, R. Chem. Ber. 1985, 118, 2548.
NMR spectrum recorded under these conditions (see Table 2) differs from that recorded in superacidic FS03H medium in that the chemical shift changes are considerably less pronounced. Most notably, the Ad for the exol endo methylene protons is reduced to 1.19 ppm. This is as expected if these shift parameters are directly related to the extent to which the carbonyl group is protonated. Childs et al. have generated complexes involving 2,3homotropone and Lewis acids having different chargestabilizing abilities and shown the chemical shifts of the methylene protons to be directly linked to the donor properties of the 0-LA- s ~ b s t i t u e n t .The ~ ~ greater acid strength of FS03H relative to CF3COOH would shift the equilibrium more extensively in the direction of 7a. Studies Aimed at Generation of the Parent Dihomotropylium Cation. Hydroxyl-substituted cations such as 33 and 34 are not the most ideal models for the assessment of delocalization because a substantial component of the positive charge resides on their oxygenated carbon atoms. For this reason, independent evaluation of the parent cations is often sought in order to gain a more rigorous appreciation of the electronic state of affairs unperturbed by substitution. In an effort to investigate the relative merits of 7b and 8b, 6 was reduced with sodium borohydride in the presence of cerium trichloride to give 35 (Scheme 4). This sensitive dienol was immediately dissolved in acetic anhydride containing 25% of glacial acetic acid and treated with a catalytic quantity of p-toluenesulfonic acid.36 The ensuing reaction did not give the acetate of 35 but resulted instead in the formation of a 30:70 mixture of the exo- and endo-bicyclo[4.2.1lnona-2,4-dien7-yl acetates 36a and 37a. Major component 37a was easily identified by means of its 300 MHz lH NMR spectrum. The signals from the admixed 36a were too weak and overlapping to be definitive. However, once saponification to generate the respective alcohols had been accomplished, the identity of 36b was clearly revealed.37 Additional confirmation of these structural assignments was realized by manganese dioxide oxidation of the combined 36b/37b sample exclusively to 38. This ketone had earlier been prepared in several laborat~ries.~~-~~ (36) (a) Babler, J. H.; Olsen, D. 0. Tetrahedron Lett. 1974, 351. (b) Babler, J. H.; Coghlan, M. J.; Feng, M.; Fries, P. J . Org. Chem. 1979,
44, 1716.
(37)Spessard, G . 0.;Hardgrove, G. L., Jr.; Mclntyre, D. K.; Townsend, D. J.; Milleville, G. S. J . Org. Chem. 1979, 44, 2034. (38) Masamune, S.; Kim, C. U.;Wilson, K. E.; Spessard, G. 0.; Georghiou, P. E.; Bates, G. S.J . Am. Chem. SOC.1975, 97, 3512. (39) Aumann, R.; Knecht, J. Chem. Ber. 1978, 111, 3927.
J. Org. Chem., Vol.59,No.19, 1994 5705
Preparation of [4,5]Dihomotropone
The driving force that underlies the bicyclo[5.1.11nonadienyl to bicyclo[4.2.llnonadien-7-ylrearrangement is necessarily thermodynamic in origin. Although this may appear unusual at first glance since 7b presumably enjoys pentadienylic delocalization and 39 can only profit from homoconjugation, the 1,a-alkyl shift that leads to 39 is accompanied by considerable strain release.
+ 39
Summary. The first dihomotropone has been synthesized. The preparative aspects of this study reflect the workability of a regioselective ring expansion involving a functionalized acyloin. The two key elements of the successful scheme were the use of ethyl diazoacetate as a reagent of choice for arrival at a suitably hctionalized bicyclo[5.l.llnonan-4-oneand the serviceability of the Garbisch procedure for introduction of the double bonds. With arrival at 6, it proved possible to demonstrate by W and NMR spectroscopy that this dienone is not measurably polarized in its ground state. Nor does its relatively unstable protonated form give evidence of dihomoaromatic character. The parent cation is particularly sensitive to Wagner-Meerwein rearrangement. Thus, by all indications, 6 can be looked upon as a dienone that is not electronically different from simpler monocyclic analogues. The structural building block represented by 6 is consequently suitable for the assessment of possible electronic perturbations brought on by relay orbital interactions of the type anticipated for 4 or 5. These issues are considered in the ensuing paper.40
Experimental Section Melting points are uncorrected. Infrared spectra were recorded on a Perkin-Elmer Model 1320 spectrometer. lH NMR spectra were recorded at 300 MHz and 13C spectra at 75 MHz on a Bruker AC-300 instrument. Mass spectra were recorded on a Kratos MS-30 spectrometer at The Ohio State University Chemical Instrument Center. Elemental analyses were performed by Scandinavian Microanalytical Laboratory, Herlev, Denmark. All separations were carried out under flash chromatography conditions on Fluka silica gel H. The organic extracts were dried over anhydrous magnesium s d fate. Solvents were reagent grade and in many cases dried prior to use. cis-1,3-CyclobutanediaceticAcid. High quality diacid could be routinely prepared by the saponification of highly purified 913 as follows. To a warm (60 "C) solution of KOH (680 mg) in water (1.2 mL) was added dropwise during 5 min 600 mg (3.0 mmol) of neat diester. The reaction mixture was refluxed for 20 h and concentrated to dryness. The residual solid was taken up in water (4 mL), acidified with concentrated sulfuric acid (850 mg) in water (2 mL), refluxed for 1 h, cooled, and filtered to separate a white solid. This material and the concentrated filtrate were stirred in ether (100 mL), filtered, evaporated, and dried to give 516 mg (100%)of the diacid as white plates, mp 154.5-155.5 "C (from water): IR (KBr, cm-') 3300, 2850, 1680; 'H NMR (300 MHz, CD3COCDs) 6 2.602.40 (m, 2 H), 2.40-2.25(m, 6 H), 1.5-1.4 (m, 2 H); I3C NMR (75 MHz, CD3COCD3) ppm 173.5, 41.2, 35.0, 29.1; MS m l z (M+ - HzO) calcd 154.0630, obsd 154.0640. Anal. Calcd for C8H1204: C, 55.81; H, 7.02. Found: C, 56.11; H, 7.05. Dimethyl cis-1,3-Cyclobutanedipropionate(10). The preceding diacid (357 mg, 2.07 mmol) was suspended in dry (40) Paquette, L. A.; Watson, T. J. J. Org. Chem., following paper
in this issue.
benzene (15 qL), treated with thionyl chloride (1.5 mL) and pyridine (2 drops) via cannula, stirred at rt for 12 h, and refluxed for 90 min. The homogeneous orange solution was concentrated on a rotary evaporator, evacuated at 0.3 Torr for several hours, dissolved in dry benzene (15 mL), and transferred during 30 min into a cold (-10 "C) dry reaction flask containing diazomethane ('25 mmol in 80 mL of ether) by filtration through a cotton plug under nitrogen. After the cessation of gas evolution, the reaction mixture was stirred at 0 "C for 2 h and at 10 "C overnight prior to concentration to a volume of approximately 25 mL. The insolubles were separated by filtration and the filtrate was concentrated, dissolved in absolute ethanol (20 mL), and treated portionwise with batches of silver acetate (0.1 g) and anhydrous triethylamine (2 mL). At this point, the mixture was refluxed for 45 min, cooled, filtered through a Celite plug, concentrated, and directly chromatographed on silica gel (elution with 4%ethyl acetate in petroleum ether). There was obtained 300 mg (63%) of 10 as a colorless oil: IR (neat, cm-l) 1740, 1440; lH NMR (300 MHz, CDC13) 6 3.63 (8, 6 H), 2.18 (t, J = 7.6 Hz, 4 H), 2.20-1.95 (m, 4 H), 1.64 (q, J = 7.5 Hz, 4 H), 1.16 (qd, J = 8.8, 1.9 Hz, 2 H); 13C NMR (75 MHz, CDC13) ppm 174.0, 51.3, 34.1, 32.4, 31.9, 31.3; MS m l z (M+) calcd 228.1361, obsd 228.1347.
Anal. Calcd for C12H2004: C, 63.14; H, 8.83. Found: C, 62.90; H, 8.84.
cis-1,s-Cyclobutanedipropionic Acid (12). A magnetically stirred solution of 10 (220 mg, 0.96 mmol) in absolute methanol (0.4mL) was treated at rt with KOH (0.4g) dissolved in water (1 mL), stirred for 20 h, concentrated to dryness, and processed in the manner described above. The resultant white solid was dried overnight in vacuo at 80 "C over P z O t~o give 192 mg (99%)of 12 as colorless flakes, mp 112.0-112.5 "C (from water): IR (KBr, cm-l) 3300, 2860, 1695, 1430, 1415, 1305, 1240, 1205; 'H NMR (300 MHz, CD3COCD3) 6 2.402.05 (m, 4 H), 2.17 (t, J = 7.6 Hz, 4 H), 1.62 (q, J = 7.4 Hz, 4 H), 1.28-1.12 (m, 2 H); 13CNMR (75 MHz, CD3COCD3) ppm 174.8, 34.7, 33.2, 32.1, 32.0; MS m l z (M+ - HzO) calcd 182.0943, obsd 182.0943. Anal. Calcd for C1&1604: C, 59.98; H, 8.05. Found: C, 59.77; H, 7.99. Bicyclo[5.1.l]nonane-3,5-dione (16). Into a dry threenecked flask was placed 36.4 mL of diethylzinc (1.1 M in followed by unpurified 1413 (8 mmol at a toluene, 40 "01) maximum) dissolved in dry ether (15 mL). This solution was stirred in a water bath as a solution of diiodomethane (12.34 g, 46 mmol) in dry ether (5 mL) was introduced dropwise during 15 min under Nz. The reaction mixture was stored at 35 "C for 3 h, cooled to 0 "C, and quenched with cold saturated NH4C1 solution (3 mL). The insoluble white solid was separated by filtration and washed thoroughly with ether (3 x 30 mL). The combined filtrates were washed with cold saturated m C 1 solution (20 mL) and brine (20 mL) prior to drying and filtration. There remained 2.46 g of 15 as a thick yellow oil. Ferric chloride was dried by overnight reflux in thionyl chloride and evacuation to 1 mm at 100 "C for 24 h. A 2.88 g sample of this material was slurried under Nz with anhydrous DMF (12 mL, freshly distilled from CaH2) at 20 "C for 15 min before being treated with the 15 prepared above dissolved in anhydrous THF (8 mL). After 3 h of stirring at rt, the dark orange solution was treated with 10% HC1 (30 mL) and extracted with CHC13 (5 x 25 mL). The combined organic extracts were washed with 10% HC1 (20 mL), a saturated NaHC03 solution (20 mL), and brine (20 mL) prior to drying and concentration. The residual DMF was distilled under reduced pressure and the residue was purified by silica gel chromatography (elution with 30%ethyl acetate in petroleum ether) to give 670 mg (55%)of 16 as a colorless crystalline solid, mp 98-99 "C (from ether): IR (KBr, cm-l) 1710,1650,1345, 1280, 1240, 1090; 'H NMR (300 MHz, CDC13) 6 3.85 (9, 2 HI, 2.85-2.60 (m, 6 H), 1.85-1.70 (m, 4 H); 13C NMR (75 MHz, CDC13) ppm 202.2, 59.3, 49.6, 29.4, 27.5; MS m / z (M+)calcd 152.0837, obsd 152.0819. Anal. Calcd for CgH1202: C, 71.03; H, 7.95. Found: C, 70.68; H, 7.95.
5706
Paquette et al.
J. Org. Chem., Vol. 59, No. 19, 1994
Bicyclo[5.l.l]nonane-3,4,5-trione4-(Propylenedithioketal) (17). A mixture of 16 (152.2mg, 1.0mmol), trimeth'~ acetate ylene dithiotosylate (477mg, 1.145m m ~ l ) ,potassium (723 mg), and absolute ethanol (8 mL) was refluxed for 24 h, concentrated to dryness, and partitioned between water (10 mL) and CHzClz (15 mL). The water layer was further extracted with CHzClz (4x 15 mL), and the combined organic phases were dried and concentrated. Chromatography of the residue on silica gel (elution with 10% ethyl acetate in petroleum ether) afforded 17 (231mg, 90%) as a colorlesscrystalline solid, mp 151-151.5 "C (from ether-petroleum ether): IR (KBr, cm-I) 1680,1440,1425,1410,1265,1165,1125,1085; lH NMR (300MHz, CDCb) 6 2.92 (t, J = 5.2 Hz, 4 H), 2.74 (br s, 4 H), 2.56 (br s, 4 H), 1.95-1.85 (m, 2 H), 1.48 (br d, J = 8.5 Hz, 2 H); 13CNMR (75 MHz, CDCl3) ppm 199.0,69.4, 43.3,30.3,28.8, 27.2,24.2;MS m l z (M+)calcd 256.0592,obsd 256.0572.
Anal. Calcd for C12H1602S~:C, 56.22;H, 6.29. Found C, 56.06;H, 6.34. Hydride Reduction of 17. To a cold (-15 "C), stirred suspension of LiAlH4 (68 mg) in dry THF (3 mL) was added dropwise under Nz a solution of 17 (150 mg, 0.585 mmol) in the same solvent ( 5 mL). After 4 h at this temperature, saturated NazS04 solution (1.5mL) was introduced and the suspension was allowed to warm to rt. The insoluble white solid was separated by filtration and washed extensively with THF (6 x 10 mL). The combined filtrates were concentrated and the residue was chromatographed on silica gel (elution with 40% ethyl acetate in petroleum ether) to give 117 mg of a 2:1 mixture of 18 and 20 and 21 mg (14%) of 21. Recrystallization of the mixture from ether-petroleum ether afforded 76 mg (50%) of pure 18. In a similar reaction carried out at -78 "C, it proved possible to isolate 20 in pure condition from the mother liquors of the first crystallization. For 18: colorless crystals, mp 147.5-148.5 "C; IR (KBr, cm-') 3200,1470,1440,1410,1300,1290,1085,1075,890; 'H NMR (300MHz, CDCl3) 6 4.43 (t,J = 4.8Hz, 2 H), 3.74 (br s, 2 H), 2.93-2.88 (m, 4 H), 2.55-2.48 (m, 4 H), 2.46-2.32 (m, 3 H), 2.30-2.26 (m, 2 H), 2.03-1.95 (m, 2 H), 1.87-1.80(m, 1 H); 13CNMR (75MHz, CDCls) ppm 77.6,64.1,36.4,30.3,29.1, 29.0,26.8, 25.5,23.9; MS m l z (M+) calcd 260.0905,obsd 260.0933. Anal. Calcd for C ~ Z H ~ ~ O C,~55.35; S Z : H, 7.74. Found: C, 55.16;H, 7.74. For 20 colorless crystals, mp 68.5-69.5 "C; 'H NMR (300 MHz, CDCl3) 6 4.17 (9, 1 H), 3.55 (t, J = 6.7Hz, 2 H), 3.253.16 (m, 2 H), 2.72 (d, J = 7.3 Hz, 2 H), 2.62-2.54 (m, 3 HI, 2.35-2.19 (m, 3 H), 2.14-1.94 (m, 2 H), 1.67-1.57 (m, 2 H), 1.46 (8, 1 H), 1.38-1.30 (m, 2 H); 13CNMR (75MHz, CDCl3) ppm201.8,61.1,47.4,46.9,40.2,35.0(2C),29.4,28.3,26.2(2 C), 25.2;MS m lz (M+)calcd 260.0905,obsd 260.0918. For 21: colorless oil; 'H NMR (300MHz, CDCl3) 6 3.86 (d, J = 6.2Hz, 1 H), 3.80-3.74 (m, 1 H), 3.54(t, J = 6.7Hz, 2 H), 2.95-2.85 (m, 2 H), 2.80-2.68 (m, 2 H), 2.40-2.15 (m, 4 H), 2.10-1.75 (series of m, 4 H), 1.68-1.58 (m, 4 H), 1.37-1.25 (m, 2 H); I3CNMR (75MHz, CDCL) ppm 71.1,61.2,52.7,41.5, 40.3,35.4,34.9,29.4,29.2,28.6,28.1,25.7; MS mlz (M+)calcd 262.1061,obsd 262.1058.
4Bromobicyclo[4,l.l]octan-3-one.A cold (0 "C),nitrogenblanketed solution of methanesulfonyl chloride (7.0mL) in anhydrous CHzClz (10 mL) was treated sequentially with anhydrous pyridine (9.7mL) and 2313(3.0g, 21.4 mmol) in dry CHzClz (10mL). The reaction mixture was stirred at 0 "C for 3 days, quenched with water (100mL), and neutralized with 10% HC1. The product was extracted into CHzClz (3 x 100 mL), washed with saturated NaHC03 solution (100 mL), dried, and evaporated. "here was obtained 3.0 g (64%) of 24 as a colorless crystalline solid, mp 112-113 "C (from CHzCLpetroleum ether): IR (KBr, cm-l) 1720,1360,1175,995,985, 970,855; 'H NMR (300MHz, CDCl3) 6 5.61 (dd, J = 7.4,10.6 Hz, 1 H), 3.22 (8, 3 H), 2.76-2.47 (m, 6 H), 2.50-2.35 (m, 1 H), 1.90-1.72(m,2H), 1.23-1.18(m,1H);'3CNMR(75MHz, CDCla)ppm205.7,82.7,46.3,39.4,35.0,31.5, 30.3,29.1,28.3; MS mlz (M+)calcd 218.0613,obsd 218.0668. Anal. Calcd for CgH1404S: C, 49.52;H, 6.46. Found: C, 49.64;H, 6.58.
A mixture of 24 (1.0g, 4.6 mmol), anhydrous LiBr (2.3g), and acetone (25 mL) was refluxed under N2 for 4 h, cooled to rt with stirring during 24 h, evaporated, and diluted with water (20mL). The product was extracted with CHzClz (3 x 20 mL), the combined organic layers were dried and concentrated, and the residue was subjected to column chromatography on silica gel (elution with 25% ethyl acetate in petroleum ether) to furnish 0.50 g (61%) of the a-bromo ketone as a colorless oil: IR (neat, cm-') 1720;'H NMR (300MHz, CDC13) 6 4.96 (dd, J = 6.8,10.9 Hz, 1 H), 2.85-2.75 (m, 1 H), 2.702.39 (m, 6 H), 1.99 (td, J = 1.5,7.5 Hz, 1 H), 1.80-1.69(m, 1 H), 1.30-1.20 (m, 1 H); 13CNMR (75MHz, CDC13) ppm 203.8, 54.6,46.4,39.9,33.4,31.2,28.8, 28.5;MS mlz (M+) calcd 201.9993,obsd 202.0028. Bicyclo[4.1.1.loct-3-en-4-one(25). A slurry of 24 (322mg, 1.52 mmol), anhydrous LiBr (1.32 g, 15.2 mmol), lithium carbonate (1.19g, 15.65 mmol), and dry DMSO (5 mL) was stirred for 3 h at 70 "C and subsequently for 2.5 h at 100 "C under N2. After cooling, the mixture was partitioned between water (10 mL) and ether (20 mL). Insoluble material was removed by filtration and washed thoroughly with ether. The aqueous phase was extracted with ether (3 x 20 mL), and the combined organic solutions were washed with water (3 x 10 mL), dried, and concentrated. Bulb-to-bulb distillation of the residue gave 84 mg (43%) of 25,bp 35-40 "C (1.5Torr), as a colorless solid, mp 38.5-39.5 "C: 'H NMR (300MHz, CDC13) 6 7.18 (dd, J = 8.7,11.5 Hz, 1 H), 6.00(d, J = 11.5 Hz, 1 H), 3.05-2.85 (m, 1 H), 2.85-2.65 (m, 5 H), 1.75-1.55 (m, 2 H); I3C NMR (75MHz, CDCls) ppm 156.1,131.1,46.6,35.6,34.4, 30.3 (carbonyl peak not seen); MS mlz (M+)calcd 122.0732, obsd 122.0761. Ethyl 3-Bromo-4-oxobicyclo[5.l.l]nonane-5-carboxylate (27). Bromo ketone 24 (2.48g, 12.3 mmol) dissolved in CHzClz (50 mL) was cooled to 0 "C under Nz, mixed with freshly distilled boron trifluoride etherate (7.5mL), and treated dropwise during 1 h with a solution of ethyl diazoacetate (6.4 mL) in CHzClz (10mL). The mixture was stirred at 0 "C for 30 min and at rt overnight, quenched with water (25mL), and stirred for 1 h. After dilution with more CHzClz, the organic layer was separated, dried, and evaporated. The residue was purified by silica gel chromatography (elution with 10% ethyl acetate in petroleum ether) to give 3.01 g (88%) of 27 as a colorless oil: IR (neat, cm-l) 1745,1715,1650;MS m l z (M+) calcd 288.0361,obsd 288.0373. Bicyclo[5.1.l]nonan-4-one(13). A dry 1 L flask equipped with a magnetic stirring bar was charged with 27 (4.29g, 14.8 mmol), ether (720 mL), acetic acid (72.1mL), and zinc dust (35.7g). This mixture was stirred under Nz at rt for 4 h and filtered. The solid was washed with water (8 x 200 mL),dried, and concentrated to give 2.69 g (87%) of the ,%keto ester, which can be directly utilized. Purification for analysis was achieved by flash column chromatography on silica gel (elution with 5% ethyl acetate in petroleum ether) t o give the product as a colorless oil: IR (neat, cm-') 3360,1745,1705,1640;lH NMR (300 MHz, CDC13) 6 4.12 (9, J = 7.1 Hz), 3.97 (dd, J = 6.8, 10.3 Hz), 2.85-2.65 (m), 2.65-2.30 (m), 2.30-2.15 (m), 2.151.90 (m), 1.80-1.65 (m), 1.55-1.40 (m), 1.21 (t, J = 7.1 Hz); 55.4,40.0, 13CNMR (75MHz, CDC13) ppm 214.2,170.2,61.0, 35.6,32.3, 30.3, 30.2,28.8,27.7, 14.0;MS m l z (M+) calcd 210.1256,obsd 210.1237. Anal. Calcd for C12H1803: C, 68.54;H, 8.63. Found C, 68.29;H, 8.80. A solution of the p-keto ester (2.69g, 12.8mmol) in acetone (25 mL) was added t o 5 M HC1 (25 mL), and the resulting soltuion was heated to reflux for 4 h, cooled, and evaporated. The resulting aqueous solution was extracted with CHzClz (3 x 25 mL), and the combined organic layers were dried and evaporated to give 1.65 g (82%) of 13. An analytically pure sample of the colorless solid, mp 89-91 "C, was obtained by bulb-to-bulb distillation and preparative GC: IR (KBr, cm-l) 1700;'H NMR (300MHz, CDCl3) 6 2.63(t,J = 6.1 Hz, 4 HI, 2.59-2.37 (m,4 H), 2.10-1.90 (m, 4 H), 1.70-1.55 (m, 2 HI; 13C NMR (75 MHz, CDCl3) ppm 220.3,40.1,32.4,30.1,28.8. Anal. Calcd for CloHleO4: C, 59.98;H, 8.05. Found: C, 59.71;H,7.99.
Preparation of [4,5]Dihomotropone
Dibromo Ketal 28. A solution of 13 (500 mg, 3.6 mmol), ethylene glycol (4 mL), andp-tohenesulfonic acid (1crystal) in benzene (10 mL) was refluxed under a Dean-Stark trap for 24 h. After cooling, solid Na~C03was added to achieve neutralization, and the benzene was evaporated. The residue was extracted with pentane (3 x 50 mL) and then with ether (50 mL). The combined organic layers were dried and evaporated to leave 520 mg (80%)of the ethylene ketal. Purification for analysis was achieved by preparative GC leading to a colorless oil: IR (neat, cm-') 1448,1380, 1290,1220; 'H NMR (300 MHz, CDC13) 6 3.94 (8, 4 H), 2.44 (m, 2 H), 2.32 (m, 2 H), 1.98 (m, 4 H), 1.70 (m, 6 H); I3C NMR (75 MHz, CDCl3) ppm 64.0, 34.5,30.9,30.3,28.5,27.6;MS m l z (M+)calcd 182.1307, obsd 182.1315. Anal. Calcd for C11HlaOZ: C, 72.49; H, 9.96. Found: C, 72.09; H, 9.90. A solution of the above ketal (124.6 mg, 0.68 mmol) in ethylene glycol (1.5 mL) was cooled to 0 "C under Nz, treated dropwise with bromine (1mL) during 15 min, and warmed to' rt. After 20 min, the mixture was poured into a stirred slurry of Na2C03 (450 mg) in pentane (10 mL). ARer the orange color had faded, the mixture was washed with water (2 x 10 mL), dried, and evaporated. Purification by silica gel chromatography (elution with 10% ethyl acetate in petroleum ether) afforded 190 mg (82%) of 28 as colorless crystals, mp 107108 "C (from ethyl acetate-petroleum ether): 'H NMR (300 MHz, CDCl3) 6 4.98 (dd, J = 10.0, 1.7 Hz, 2 H), 4.40-4.25 (m, 4 H), 2.65-2.30 (m, 8 H), 1.80 (br d, J = 12 Hz, 2 H); 13CNMR (75 MHz, CDCl3) ppm 110.9, 68.2, 58.1, 38.2, 32.6, 28.8; MS m l z (M+)calcd 339.9496, obsd 339.9542. Anal. Calcd for CllH18Br202: C, 38.85; H, 4.74. Found: C, 39.17; H, 4.87. Bicyclo[5.1.1]nona-2,5-dien-4-one (6). A solution of 28 (146 mg, 0.63 mmol) and potassium tert-butoxide (300 mg) in dry DMSO ( 5 mL) was stirred at 50 "C for 15 h, cooled, poured into brine (10 mL), and extracted with pentane (2 x 20 mL). The combined organic extracts were dried and concentrated t o give 50 mg (65%) of the diene ketal, which was directly shaken with 3% sulfuric acid (2 mL) for 5 min, immediately extracted with ether (2 x 10 mL), dried, and concentrated. Silica gel chromatography (elution with 10%ethyl acetate in petroleum ether) gave 18 mg (47%) of 6, a colorless oil that was further purified by preparative GC: IR (neat, cm-l) 1653, 1623; 'H NMR (300 MHz, CDCL) 6 6.61 (dd, J = 12.8,8.6 Hz, 2 HI, 6.00 (dd, J =13.1, 0.8 Hz, 2 H), 3.05-2.75 (m, 4H), 1.83 (dm, J = 12.7 Hz, 2 H); 13CNMR (75 MHz, CDC13) ppm 195.0, 146.8, 130.8, 31.7, 30.1; MS m l z (M+) calcd 134.0731, obsd 134.0769. Bicyclo[5.1.1]nona-2,5-dien-4-ol (35). To 1 mL of a 0.4 M solution of CeC13.7H20 in methanol were added sequentially 10 mg of 6 and 5 mg of sodium borohydride. f i r 5 min, water (1mL) was added, the product was extracted into ether (3 x 2 mL), and the combined organic phases were dried and evaporated to produce 36 as a white solid that yellowed quickly at rt. This material was therefore directly acetylated. Acetylative Rearrangement of 35. The alcohol sample from above was dissolved in 1mL of a 4:l acetic anhydrideglacial acetic acid mixture. p-Toluenesulfonicacid (1mg) was
J . Org. Chem., Vol. 59,No. 19, 1994 6707 introduced, and the mixture was stirred at rt for 24 h, evaporated in vacuo, and diluted with ether (3 mL). The ethereal solution was washed with water (3 x 3 mL) and 5% NazC03 solution prior to drying and solvent evaporation. There was isolated 10 mg of a 30:70 mixture of 36a and 37a as a colorless mobile oil. For 37a: 'H NMR (300 MHz, CDCk) 6 6.08 (br dd, J = 11,8 Hz, 1H), 5.88 (dd, J = 12, 7 Hz, 1H), I), (ddd, J = 9, 7, 7 Hz, 1H), 3.02 (br 5.72-5.62 (m, 2 €5.06 ddd, J = 7,7,6 Hz, 1H), 2.61 (br dddd, J = 9, 8, 7, 3 Hz, 1H), 2.28 (ddd, J = 14, 9, 9 Hz, 1H), 2.12 (br ddd, J = 13, 7, 6 Hz, 1 H), 2.03 (s, 3 H), 1.92 (br d, J = 13 Hz, 1 H), 1.85 (dddd, J = 14, 7, 3, 2 Hz, 1H); 13CNMR (75 MHz, CDC13) ppm 171.2, 139.8, 132,7,127.0, 123.786.4,42.5,40.2,35.0,29.7,21.0; GCMS m l z (M+)calcd 178.10, obsd 178.20. The acetate mixture (10 mg) was dissolved in 50% aqueous methwol (1 mL) to which KOH (50 mg) was added. The mixture was stirred at rt for 1h, concentrated in vacuo, and triturated with ether (3 x 3 mL). The combined organic layers were'dried and evaporated to leave 6 mg of a 30:70 mixture of 36b and 371, as a colorless mobile oil. For 36b: 'H NMR (300 MHz, CDCl3) 6 6.11 (br dd, J = 11, 8.5 Hz, 1H), 5.92 (br dd, J = 12, 7.5 Hz, 1H), 5.75 (ddd, J = 12, 7, 1Hz, 1H), 5.64 (dd, J = 11,7 Hz, 1H), 4.34 (br d, J = 6 Hz, 1H), 2.84 (br ddd, J = 8.5,8.5, 7 Hz, 1H), 2.59 (br dd, J = 7.5, 6 Hz, 1 H), 2.39 (br ddd, J = 12, 7, 6 Hz, 1 H), 2.23 (br dd, J = 14.5, 6 Hz, 1H), 1.88(br dd, J = 14.5, 8.5 Hz, 1 H), 1.67 (br d, J = 12 Hz, 1H). For 37b: 1H NMR (300 MHz, CDCl3) 6 6.12 (br dd, J = 11, 8 Hz, 1 H), 6.03 (dd, J = 12, 7 Hz, 1H), 5.79 (br dd, J = 12, 7 Hz, 1H), 5.66 (dd, J = 11,7 Hz, 1H),4.40 (ddd, J = 9, 7, 7 Hz, 1H), 2.92 (br ddd, J = 7, 7,6 Hz, 1H), 2.58 (br dddd, J = 9, 8, 7, 3 Hz, 1 H), 2.26 (ddd, J = 14, 9, 9 Hz, 1H), 2.10 (br ddd, J = 13, 7, 6 Hz, 1 H), 1.89 (br d, J = 13 Hz, 1H), 1.68 (dddd, J = 14, 7, 3, 2 Hz, 1 H); 13C NMR (75 MHz, CDCl3) MS ppm 140.9,132.6,128.6,123.1,85.1,44.8,44.7,35.8,29.7; m l z (M+)calcd 136.0888, obsd 136.0889. Bicyclo[4.2.l]nona-2,5-dien-7-one(38). The above alcohol mixture (6 mg) was placed in CDC13 (0.5 mL) and treated with activated manganese dioxide (10 mg). The mixture was stirred at rt for 24 h and filtered through a small Celite pad into an NMR tube. 38: 'H NMR (300 MHz, CDC13) 6 6.20 (br dd, J = 11, 8 Hz, 1 H), 5.96 (dd, J = 11, 7 Hz, 1 H), 5.88 (br dd, J = 11, 8 Hz, 1H), 5.82 (ddd, J = 11, 7, 1Hz, 1 H), 3.38 (br dd, J = 8,8 Hz, 1H), 2.96 (ddddd, J = 8.5, 8,6, 2.5, 1Hz, 1H),2.55 ( d d d , J = 18.5, 8.5,1Hz, 1H), 2.42 (br d , J = 18.5 Hz, 1H), 2.37 (br ddd, J = 12.5, 8, 6 Hz, 1 H), 1.83 (dd, J = 12.5,2.5 Hz, 1H); MS m l z (M+)calcd 134.0732,obsd 134.0735.
Acknowledgment. We thank the National Science Foundation for financial support and Dr. Brian Roden for performing several probe experiments. Supplementary Material Available: 300-MHz 'H and 75-MHz 1% N M R spectra of those compounds lacking combustion data (8 pages). This material is contained in libraries on microfiche, immediately follows this article in the microfilm version of the journal, and can be ordered from the ACS; see any current masthead page for ordering information.
